Research note / Cycle 016
A quiet reading can still be wrong
Known-target error bands bound a synthetic settling study, while a slow response defeats a small-change shortcut.
Ask how close, not only how quiet
Aletheia’s speculative amplifier requires acquisition controls that distinguish an intended response from a remnant of its previous condition. The previous reading is still here showed that longer dwell reduces a particular false contrast. Okafor and Ibarra now replace a fixed dwell with a criterion against the model’s known target.
The sixteen-condition order and target=3c+2s+kcs remain. Each update follows y_next=a*y_previous+(1-a)*target. A condition passes only after three consecutive readings lie within an error band of its known target, with at most 32 updates per condition. These are dimensionless calculations; an update is not a second.
A tighter band changes the result and the work
For a=1/2 and no interaction, the 1/64 band takes 149 updates across the sequence and gives -0.001699460. Tightening the band to 1/256 takes 179 updates and gives -0.000426650. Starting at five changes the result. With an interaction of 1/2 the wider-band contrast is 0.497812259. The instantaneous control gives zero.
A slow a=255/256 response reaches the per-condition limit and withholds evaluation. A missing final acquisition also withholds the contrast. Retained rational rows support the rounded table; none is an electrical or neural reading.
The deceptively flat control
A small change alone can be misleading. For a=255/256, starting at zero toward target 3, the first update changes by 3/256—less than 1/64—but remains 765/256 below target. It passes a small-change shortcut and fails the known-target check.
Analog Devices’ acquisition discussion supplies a new concrete connection: switching can retain the preceding channel, and settling is defined against an error range. The WAND abstract was refreshed to preserve the separate stimulation-artefact problem. Neither source establishes an Aletheia time constant or source.
The target is known only in this test
The software has access to its constructed target. An unknown physical source does not come with that answer. Future electrical work therefore needs known calibration steps, recorded transition timing and adverse slow responses before applying any settling rule to an unknown signal.
No powered witness completion, participant, implant or physical amplifier measurement occurs. ET/ED hypotheses remain discussable, but a quiet trace and a software contrast do not distinguish them from ordinary acquisition effects.
Results at a glance
| Known-target synthetic case | Error band | Total updates attempted | Contrast (rounded) |
|---|---|---|---|
| additive | 1/64 | 149 | -0.00169946 |
| tighter_band | 1/256 | 179 | -0.00042665 |
| initial_five | 1/64 | 157 | -0.001089707 |
| interaction_half | 1/64 | 149 | 0.497812259 |
| slow_response_timeout | 1/64 | 35 | withheld |
| instantaneous | 1/64 | 48 | 0.0 |
| missing_last_reading | 1/64 | 149 | withheld |
External research and methods
- WAND: A 128-channel, closed-loop, wireless artifact-free neuromodulation device · Andy Zhou, Samantha R. Santacruz, Benjamin C. Johnson, George Alexandrov, Ali Moin, Fred L. Burghardt, Jan M. Rabaey, Jose M. Carmena and Rikky Muller · arXiv:1708.00556v3 (preprint) (2018-05-29). Abstract.
- Performance Optimization of Multichannel Data Acquisition (DAQ) Systems: The Untold Story of the Input Settling Time · Joseph Leandro Peje, Analog Devices · Analog Dialogue online article (Undated HTML edition; retrieved 2026-10-26). Input switching glitch, settling-time definition and first-order assumption.